Trajectory Planning And Path Following Techniques Of Unmanned Surface Vehicle

Unmanned Surface Vehicle(USV)has the advantages of modularized,autonomous and intelligentization etc,besides,when assembled with different sensors and modules,it can exucate a large amout of complex assignments,therefore,USV has been widely applied in the field of ocean exploitation and military applications.The corresponding techniques of USV have gained increasingly attentions,among of which,the extent of the autonomous navigation technique is the key element to enhance the intelligence level of USV,and hence,it becomes the hot-spot issue gradually.One of the aspects of the intelligent behavior of USV is to fulfill various kinds of missions under the presence of complex marine environment,furthermore,to adjust its state according to the frequently change of ambience.This thesis focus on relevant techniques to enhance the intelligent ability of USV,and the main works of this thesis are as follows:To solve the imperections of Intelligent Wate Drops(IWD)algorithm,such as easily trapped into local optimal solution and slow convergence speed,we improved the basic IWD algorithm by introduce the global optimal solution emphasize mechanism and the global optimal solution neibourhodd expansion mechanism,and the improvd algorithm is so called NIWD algotithm.The diversity of solution can be increased to a large extent via introducing the global optimal solution neibourhodd expansion mechanism,consequently,the local optimal solution trapping can be avoided.The convergence speed can be accelerated by introducing the global optimal solution emphasize mechanism.The value convergence is derivated in theory,moreover,the perfomnace of NIWD algorithm has been tested by a serious standard benchmark functions.After that,we proposed a trajectory planning method of USV by employing the NIWD algorithm,and the simulation results show that the propsed method is efficient and resonbale.We analysis the issue of Collision Risk Incidence(CRI)evaluation,and interpret the key factors that have great influence on the CRI evaluation.We have derived the membership function corresponding to each factor,and the Analytic Hierachy Process(AHP)method is employed to assign the weight to each key factor according to the level of influence on CRI.The defination of CRI has the characteristics of ambiguous,uncertain and instantaneous,while the Evidential Reasoning(ER)theory has outstanding performance when handling the incomplete and uncertain information fusion problem,therefore we adopt the ER theory to assess CRI to provide the reasonable decision for USV to take action of collision avoidance.The specific procedure of the proposed CRI evaluation method is presented in detail and the validity of the proeposed method has been testified through a group of simulations.To solve the drawback of the exsiting real-time collision avoidance method for USV,we proposed a real-time collsion avoidance for USV based on extended Optimal Reciprical Collision Avoidance(ORCA)algorithm.The typical encouter situations when USV executing the mission on sea have been anaylized,and the relevant collision avoidance maneuver have been presented according to the rules stipulated in the Convention on the International Regulations for Preventing Collsion at Sea(COLREGs).The traditional ORCA algorithm has been extended accordingly.Firsly,the uncertain information has been dealed with by using Kalman filtering,then the COLREGs rules are treated as the constraints when computing the optimal collision avoidance velocity,lastly,the action of the obstacles are taken into consideration to make a collision prevention decision.Moreover,the detailed procedure of the propsed method is explained elaborately.The simulation results exhibit the reliability and safety of the presented method.An H ? tracking control law is designed for the non-linear system,of which the parameters are uncertain and easily affected by the external disturbance.The Fuzzy-Model-Based(FMB)system based on Parallel Distributed Compensiation(PDC)mode and Imperfect Premise Matching(IPM)mode are explained separately,and the shortcoming of PDC FMB system is declared.We adopt the Takagi-Sugeno(T-S)fuzzy method to establish the model of non-linear system,the perturbation of system paremeters and the external disturbace are also taking into account when modeling the system.We design a state observer to estimate the nonmeasurable state variables,after that,an H ? tracking control law based on IPM mode is propsed and the stability condition while fulfill the H ? tracking performace is derived,moreover,the Linear Matrix Inequalities(LMI)form of the stability contition is deduced.The effectiveness of the propsed traking method has been valiated by an inverted pendulum on a car,the simulation results show that this method has great robustness ability against the disturbace and perturbation.We analysis the kinematics and dynamic model of USV,and simpliy the model according to the USV path following problem,the three degrees of freedom of USV motion model is presented.The differential global homeomorphism transformation method is employed to transform the dynamic model of USV and the tracking system to obtain the tracking error dynamic model.Afterwards,the casecaded approach is adoped to transform the tracking control problem to the stabilization control issue of two independed error dynamic system.Subsequently,the validity of the propsed path following controller for USV is testified through simulation results.